PROJECT SUMMARY/ABSTRACT Stroke rehabilitation research has produced a number of efficacious and emerging interventions for improving mobility, balance, and upper limb function. Trial outcomes are measured at the functional capacity level, where capacity is defined as what a person is capable of doing in the structured environment of the clinic or laboratory. It has long been assumed that improvements in capacity result in improvements in real-world performance, where performance is defined as what a person actually does in daily life, outside of the clinic or laboratory. Our data from the previous award cycle show a striking difference after intervention between capacity and performance measured with wearable sensors, i.e. accelerometers. Group averages indicated some improvement in capacity and in self-reported performance, but no improvement in actual performance. At the individual level, no one showed improvements in performance, even those individuals who made substantial changes in capacity. Given that a key purpose of providing rehabilitation services post stroke is to improve performance in daily life, we have uncovered a major problem that merits investigation. This project will examine the discrepancy between capacity and performance, with the long-term goal of using the knowledge gained to develop new or modify current rehabilitation interventions that will improve performance in daily life. We will study two prospective longitudinal cohorts to address three aims. Aim 1 will seek to understand the scope of the problem within the field of neurorehabilitation. We will study a cohort of people receiving outpatient services to determine whether the discrepancy between capacity and performance is unique to upper limb interventions and/or stroke rehabilitation. Aim 2 is designed to learn when, how much, and in whom, upper limb capacity gains translate to performance gains. We will study another cohort of persons with first time stroke to map the natural trajectory of performance and its relationships to capacity and other factors over the time course of stroke recovery (within 2 weeks out to 6 months). Aim 3 takes the viewpoint that upper limb performance is a health behavior, or habit, which may be amenable to change. We will capitalize on the same cohort as Aim 2 and explore the time course of stroke survivor attitudes and barriers to performance. Data from these cohorts will transform current practice with knowledge of when and with whom motor rehabilitation interventions can improve performance in daily life. Rehabilitation research will be advanced by knowledge about capacity vs. performance, and when and how capacity improvements can translate to performance improvements. Our data will inform future research design and serve as a basis for developing and testing of performance-level interventions to improve stroke outcomes in the real world, not just in the clinic or laboratory.